Linux CoC → different perspective

mudasobwa profile image Aleksei Matiushkin ・3 min read

Unfortunately, dev.to does not [yet?] provide an ability to cite/refer to another resource, so I took the responsibility of copy-pasting this brilliant email originally sent to LKML here.

This is a must-read for everyone who thinks that Code of Conduct protects people and makes them happier. That is not always the case, and here is why.

From      Edward Cree
Subject   Re: Code of Conduct: Let's revamp it.
Date      Wed, 19 Sep 2018 07:00:26 +0100

The new Code of Conduct makes me feel threatened and uncomfortable.

No, really.  As a person with (diagnosed) Asperger's, I'm a member of,
  objectively, a marginalised minority.  Effectively (i.e. this is a massive
  oversimplification), I was born without the hard-wired circuitry for social
  interactions that is normally a part of the human brain; consequently I have
  to run a slow and inaccurate software simulation when interacting with
  'normal' people.

In nearly all the communities I participate in, this is a constantly limiting
  factor for me.  But there is one world that is blessedly free of such things:
  the world of open-source software.  It is one of the last places where my
  particular neurodiversity does not mark me out as Other, does not force
  me to carefully watch what I say and present a falsely constructed façade in
  place of my real identity.  For here, we care not for 'feelings'; either the
  code is good or it is bad, and in the latter case we say so directly and
  bluntly.  Not only does this mean that I don't have to guard my tongue when
  critiquing someone else's patch, far more importantly it means I can
  understand what's being said when my patches are criticised.

(Almost all
  of my best ideas and patches have been born out of someone telling me I'm

The Linux kernel community is a place without office politics, without subtle
  subtexts, without primate dominance dynamics.  A place where criticism can
  be gracefully accepted without having to worry that admitting to being
  wrong will lower one's status.  A place where I, and people like me, can feel
  at home, and maybe even create something of value.

And the Contributor Covenant looks very much like the camel's nose of an
  attempt to take that place, that community, away from me.  To replace it with
  an Orwellian nightmare where I must forever second-guess what is safe to say.
  (First they came for "master/slave replication", and I did not speak up
  because I was not a DBA.)

I cannot speak for my employer (hence why I am posting this from my personal
  address), but to the extent that my rôle as a contributor to the networking
  subsystem, and as co-maintainer of the sfc driver, gives me any standing in a
  personal capacity, I absolutely cannot sign up to this 'Pledge' nor accept
  the 'Responsibilities' to police the speech of others that it makes a duty of
  maintainership, and I urge the project leadership to revert its adoption.

Some elements of the Code are unobjectionable; sexual advances, for instance,
  have no place on the lkml (though they may at, say, a conference, and not
  everyone can reliably predict whether they are unwelcome), and the ability of
  kernel developers to accept constructive criticism is one of the strengths
  that has made Linux what it is.  But far too many of its provisions rely on
  ill-defined terms, and thus give those charged with interpreting those terms
  the power to destroy livelihoods.  By placing a corporate body (the LF) in
  the position of arbiter, an avenue is opened for commercial pressure to be
  applied; and the legalistic phrasing of the Code practically invites rules-
  lawyering whereby the most abusive may twist it into a weapon to further
  their abuse.

If the Code were reduced to something more like the old Code of Conflict,
  reminding people to 'be liberal in what they accept and conservative in what
  they emit', and clarifying that patch submissions should be judged by the
  code and not by any characteristics or beliefs of the submitter (I don't
  think the enumerated list of protected classes is helpful, as a legalistic
  abuser can always slip into a crack between them), I think the sting would be
  drawn.  Probably the CoConflict would make a better base from which to draft
  such a document.

(A note for the irony-challenged: where I use Progressive terms-of-art, such
  as 'marginalised', 'Other' and 'identity', in the above, I am endeavouring to
  show that this alleged push for 'inclusiveness' fails on its own terms; I am
  not accepting the theory behind those terms nor suggesting that, in
  reality, the kernel community owes me any special treatment on account of my


Posted on Sep 27 '18 by:

mudasobwa profile

Aleksei Matiushkin


NB! I am _not_ a member of #DEVCommunity. → I like: Elixir, Erlang, Ruby, R, C, COBOL. → I hate: Apple, JS, Rails, haters. → I am more functional, than object oriented.


markdown guide

The Linux kernel community is a place without office politics, without subtle subtexts, without primate dominance dynamics.


Any talk about Linux's relationship to Microsoft has always been a primate-dominance-fight. Same reason why the stock Linux kernel dumps dmesg too-fast-to-read to the terminal. And, of course, you're engaging in monkey dominance posturing by posting that sentence.

Sometimes, it can become an intra-subjective fact, like how people's belief that something is fit-for-purpose can lead to them building the tooling that makes it true (that's why C, with its many extensions and compiler optimizations, is better for OS development than anything else currently is, even if we'd be better off if Ada had won). This stuff can sometimes make my head hurt, too, and I'm not even on the spectrum.

Having Aspergers means you have trouble reading social cues. It doesn't mean you don't commit logical fallacies just as badly as the rest of us.

And, most importantly, none of those previous paragraphs are a dig at you or at LKML. Acting like a person isn't wrong, and if you feel safer there than in other places that value politeness more highly, that's cool, that's important, and that's doesn't conflict with spelling out, in so many words, common forms of discrimination that are not allowed. Just drop the "I achieved nirvana" shtick.


Yes, these are valid difficulties, but Edward is describing a nostalgia which cherry picks.

The Linux community is struggling with "some big change all of a sudden", but it didn't happen until after this technology became globally dominant. The healthy approach would have been to grow and mature as a project along with its success, in the Linux way, but because the community clung to values that don't scale, they ultimately had less control over the outcome.

Unwillingness to change at all, meant that they couldn't stay ahead of these issues. Linux could have lead the way instead of putting up a fuss and succumbing to their own success.


Off-topic, but what do you mean by

Same reason why the stock Linux kernel dumps dmesg too-fast-to-read to the terminal.

I don't see why that's a problem? If something crashes, you get the last lines right in front of you, and if you want to review it later you can, so why is this a bad thing?


Unfortunately, you did not get the message. That’s fine, just drop the “I know how” shtick.


The OP is afraid of accidentally pissing someone off, and thinks that the CoC means that he'll be banished for it.

Am I wrong?

IMHO you are absolutely wrong.

The OP is mostly insulted by the feeling that Linus (and some others) were literally harassed and enforced to do things that make them unhappy.

I am pretty sure OP had never nearly pissed someone off and had less than zero plans to do that in the future.

That's not the story Linus tells. And I think it's an insult, especially to someone who prides himself on his thick skin, to claim that the Twitter mob made him do it.

That mail says that the CoC change was made because the code of conflict had made the kernel summit less fun.

Though, thanks. You're right. I think I was missing the point.

I know what story tells Linus (I hope you somehow assumed I had it read before.) I see the highlight of the Linus’ email in:

Maybe I can get an email filter in place so at when I send email with curse-words, they just won't go out. Because hey, I'm a big believer in tools, and at least some problems going forward might be improved with simple automation.
I know when I really look “myself in the mirror” it will be clear it's
not the only change that has to happen, but hey... You can send me
suggestions in email.

I see this as a bitter irony, nearly trolling. I might be wrong, but hey, I might be right as well. And even if we just imagine the world where I am right, we’d immediately see what am I trying to express by pushing the OP and saying all that crap I am saying here.


It seems to me that it should be pretty easy to maintain high standards and quality while at the same time not engaging in mean or hurtful interactions with people. It probably makes things sound kind of technocratic, but I don't see that as a bad thing. We’re talking about kernel development after all. If you don't agree with a code submission, just say something like:

  • The following code doesn't meet our standards because of a, b, and c. Please fix these problems and re-submit.
  • This approach won't work because it will result in problems a, b, and c.
  • This code increases complexity while not actually solving the problem it purports to deal with, for the following reasons. Don't use this approach in kernel code submissions going forward.

This is very minimal. It doesn't go out of its way to be extra nice, but I think it's polite enough, no? I'm sure if Linus and the other core people behind the kernel just limited themselves to this kind of wording, there would not have been such an outcry about the code of conduct in the first place.

The problem is this kind of thing, written by Linus himself, that has become well known:

lkml.iu.edu/hypermail/linux/kernel... (contains abusive language):

Christ people. This is just sh*t...The conflict I get is due to stupid new gcc header file crap. But what makes me upset is that the crap is for completely bogus reasons... and anybody who thinks that the above is... is just incompetent and out to lunch... Really. Give me one reason why it was written in that idiotic way...So I really see no reason for this kind of complete idiotic crap... And it's a f*cking bad excuse for that braindamage...

I'll stop there. My point is: Is this kind of thing more acceptable for people with something like Asperger's? If someone with Asperger's got this kind of feedback to a code submission, would it not be even harder for them to process than for a neurotypical person?

Not everyone is cut out to wrap every sentence in warm and fuzzy language. I think that's fine. In that case, just stick to clear and neutral language. Don't launch into foul or abusive tirades. And if it's possible to explain how a person can improve in the future, do that, because it will have the practical benefit of helping them to make better contributions. I don't get why this needs to be so hard.


My point is: Is this kind of thing more acceptable for people with something like Asperger's?

Unfortunately, you did not get the message. Asperger’s has nothing to do with the real issue. It was an example, the author said many times it’s just an example—and still, every single comment here appeals to Asperger’s.

I do not have Asperger’s. And I pretty much on this email author’s side. Here is why.

it should be pretty easy to maintain high standards and quality while at the same time not engaging in mean or hurtful interactions with people

Well, maybe. That is out of scope, actually. The subject here is different: “be polite” worship came without any doubt that assaulting and harassing people who are out of this cult is fine. Being polite is given as a God's gift. Here is the truth: it is not by any mean. Despite you might think differently. We are fine with those wanting to be polite, but they are not fine with us.

Here are some examples of the same pattern from the past (including but not limited to):

In all the examples above the majority was sure it does well and the minority should just obey for their own benefit.

If you think I am exaggerating, I am not.

This is how damn bloody majority rule works.

And harassing, baiting and assaulting people like Linus or the author of this email—cannot be good no matter what the shining goal is it for.

Dictating any culture is an enforcing a slavery for others. Even while masters are sure everything is done for these stupid people goodness.


Yeah, I do think you're exaggerating, and I'd be surprised if even Edward Cree (the OP) would go as far as you are here. Power Mods who treat their mailing list like a fiefdom, and social media klans who inundate people with negativity, are not the same as the real thing. They don't have access to physical force. It's the internet. You can always leave.

Which is not to be taken as sticking up for the Twitter mob. There's a reason my Twitter account is locked.

I'm sticking up for the Code of Conduct specifically. After all, DEV has a very similar code, and you seem happy enough with it. I'm not getting kicked out for using strong language, and you're not getting kicked out for invoking Godwin's Law. I expect the LKML to be less like Tumblr and more like Lobsters, really.

you seem happy enough with it

Nah, I am not. Ask @ben : I was nearly banned twice, I was personally emailed by core team members to stop exposing my thoughts broadly and my innocent post containing no one harsh or rude word was removed because some anonymous but still way-too-tolerant people literally conveyed I am a moron insulting people.

I'm not getting kicked out for using strong language

My wild guess would be it’s only because I am known as an easy person to not to pay attention to words and focus on the meaning. If I were a bit less intelligent, you’d received all that fancy load of the tolerance (see above.)

and you're not getting kicked out for invoking Goodwin's Law

I explicitly avoided the protagonist of Godwin’s law, so nice try, but plain wrong again.

I really don't care about the specific distinction between Nazi Germany and the Spanish Inquisition. It's the same non-argument, comparing online harassment to a state-sponsored atrocity. They are both bad, but they are not anything near the same.

I'm curious about how "nearly" those near-bans were. I've had plenty of conversations with the mods (not here, not been around long enough, but in similarly strongly-moderated forums), and participated in threads that got locked. It's actually really difficult to get banned from a place like this, and I expect LKML will be the same. Lots of warnings and locks, very few bans.

You might care or don’t care about whatever you want, you might even claim that 2×2=5 and live in that particular axiomatics, but the communication is easier when people use the same meaning of words and that’s where the vocabulary comes to the scene. Godwin’s law mentions Hitler. Period.

I'm curious about how "nearly" those near-bans were.

Ad hominem arguments are great to bully the person but they rarely if never help to prove anything. It does not really matter ‘how "nearly" those near-bans were,’ what actually matters is ‘how insulting they were.’

That’s the point I advocating for and that’s IMHO the main point of the OP.


As another individual with Aspergers, I 100% understand where you are coming from with this. I feel targeted by this. It wasn't required, and was simply a target against a single individual I feel.

I feel segregated when I read this because I feel that I should not be treated any differently than anyone else, regardless of who I am. If you cannot code, don't contribute code to a system. If someone says "You need to follow this style guide", accept that you have to do that.

This whole "Code of Conduct" thing resonates with me that they are targeting Linus and only him, especially the Tweet that was put out by the person who had it implemented (I forget the name of them) saying they cannot wait for the Mass exodus of Linux.



Unfortunately, dev.to does not [yet?] provide an ability to cite/refer to another resource, so I took the responsibility of copy-pasting this brilliant email originally sent to LKML here.

That's what things like the line-leading > and the markdown standard link-out methods are for. Probabably other methods you could leverage, as well (depending how much they excised from their markdown implementation).